model/README.md

# InterFuser Model Directory

## Required Files

Place your trained InterFuser model files in this directory:

1. **`interfuser_model.pth`** - The main model weights file
2. **`config.json`** (optional) - Model configuration file

## Model Format

The model should be a PyTorch state dict saved with:
```python
torch.save(model.state_dict(), 'interfuser_model.pth')
```

## Loading in Code

The model is loaded in `model_definition.py`:
```python
model = InterFuserModel()
model.load_state_dict(torch.load('model/interfuser_model.pth', map_location='cpu'))
```

## Note

- The current implementation uses a dummy model for testing
- Replace with your actual trained InterFuser weights
- Ensure the model architecture matches the one defined in `model_definition.py`